Article Text

Download PDFPDF
Angiopoietin 2 expression in the cornea and its control of corneal neovascularisation
  1. Giulio Ferrari1,2,
  2. Chiara Giacomini1,
  3. Fabio Bignami1,
  4. Davide Moi3,4,
  5. Anna Ranghetti3,
  6. Claudio Doglioni5,
  7. Luigi Naldini3,5,
  8. Paolo Rama1,2,
  9. Roberta Mazzieri3,4
  1. 1Eye Repair Laboratory, Department of Neurosciences, IRCCS San Raffaele Scientific Institute, Milan, Italy
  2. 2Cornea and Ocular Surface Unit, IRCCS San Raffaele Scientific Institute, Milan, Italy
  3. 3Division of Regenerative Medicine, Stem Cells and Gene Therapy, San Raffaele Telethon Institute for Gene Therapy (TIGET), IRCCS San Raffaele Scientific Institute, Milan, Italy
  4. 4The University of Queensland Diamantina Institute, Translational Research Institute, Brisbane, Queensland, Australia
  5. 5Vita Salute San Raffaele University, Milan, Italy
  1. Correspondence to Dr Giulio Ferrari, Eye Repair Laboratory, Department of Neurosciences, IRCCS San Raffaele Scientific Institute, Via Olgettina 60, Milan 20132, Italy; ferrari.giulio{at}


Purpose To define proangiogenic angiopoietin 2 (ANG2) expression and role(s) in human and mouse vascularised corneas. Further, to evaluate the effect of ANG2 inhibition on corneal neovascularisation (CNV).

Methods CNV was induced in FVB mice by means of intrastromal suture placement. One group of animals was sacrificed 10 days later; corneas were immunostained for ANG2 and compared with (i) mouse non-vascularised corneas and (ii) human vascularised and non-vascularised corneas. A second group of CNV animals was treated systemically with an anti-ANG2 antibody. After 10 days, the corneas were whole-mounted, stained for CD31 and LYVE1 and lymphatic/blood vessels quantified. In another set of experiments, the corneal basal Bowman membrane was either (i) removed or (ii) left in place. After 2 or 10 days the corneas were removed and immunostained for collagen IV, ANG2, CD31, LYVE1, CD11b and MRC1 markers.

Results In human beings and mice, ANG2 is expressed only in the epithelium, and, mildly, in the endothelium, of the avascular cornea. Instead, it is expressed in the epithelium, endothelium and stroma of vascularised corneas. Disruption of the Bowman membrane is associated with a significant increase of (i) ANG2 stromal expression and (ii) proangiogenic macrophage infiltration in the corneal stroma. Finally, blocking ANG2 significantly reduced hemangiogenesis, lymphangiogenesis and macrophage infiltration.

Conclusions Balancing proper healing and good vision is crucial in the cornea, constantly exposed to potential injuries. In this paper, we suggest the existence of a mechanism regulating the onset of inflammation (and associated CNV) depending on injury severity.

  • Cornea
  • Neovascularisation
  • Inflammation
  • Angiogenesis
  • Ocular surface

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

  • At a glance
    Keith Barton Jost Jonas James Chodosh